Can temperature be used to tune the selectivity of membrane ion-selective electrodes?

Article abstract of DOI:10.1021/ac902867d

The selectivity coefficients, K_IJ^pot, of ion-selective electrodes (ISEs) have been fundamentally related to thermodynamic parameters, but yet, the effect of temperature on K_IJ^pot has not been studied. We describe a new approach to fine-tune the selectivity of ISEs based on the effect of temperature on selectivity coefficients. This effect was quantified for ion exchangers as well as neutral carrier ionophores. Potassium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate was used as an ion exchanger in a poly(vinyl chloride) membrane. The corresponding electrode showed an inverse relationship between log K_Na^pot+,_K+ and temperature. Additionally, sodium-selective electrodes using monensin, monensin methyl ester (MME), and monensin decyl ester (MDE) were evaluated as models of electrodes based on neutral carriers. The electrode based on monensin showed an enhancement in the selectivity for sodium over potassium by half an order of magnitude as the temperature increased from 20 to 50 °C, while the electrodes based on MME or MDE showed a very small change in selectivity. This can be explained in light of changes in the formation constants between the ions and ionophores with temperature. The theory of the effect of temperature on selectivity for both ion exchangers and neutral carrier ionophores is also discussed.

Full text of DOI:10.1021/ac902867d

Anal. Chem. 2010, 82, 3622–3628
Can Temperature Be Used To Tune the Selectivity
of Membrane Ion-Selective Electrodes?
Elsayed M. Zahran,† Vasileios Gavalas,†,‡ Manuel Valiente,§ and Leonidas G. Bachas*,†
Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506, and Departament de Qu´ımica,
Universitat Auto`noma de Barcelona, E-08193 Bellaterra, Barcelona, Spain
pot
The selectivity coefficients, K , of ion-selective elec-
selectivity properties,1-3 optimization of the concentration of
lipophilic anionic and cationic sites in the membrane,4-7 adjust-
ment of the internal solution composition and pH,8 use of
asymmetric membrane configurations,9 and control of ion fluxes
between the membrane and the solution phase by pulsed chro-
nopotentiometry.10
IJ
trodes (ISEs) have been fundamentally related to
thermodynamic parameters, but yet, the effect of tem-
perature on K has not been studied. We describe a
new approach to fine-tune the selectivity of ISEs based
on the effect of temperature on selectivity coefficients.
This effect was quantified for ion exchangers as well
as neutral carrier ionophores. Potassium tetrakis[3,5-
bis(trifluoromethyl)phenyl]borate was used as an ion
exchanger in a poly(vinyl chloride) membrane. The
pot
IJ
The true thermodynamic meaning of the Nicolsky-Eisenman
11
pot
empirical selectivity coefficient, K
,
,12 has been studied exten-
IJ
sively over the past two decades.13-16 Different mathematical
expressions have been derived for the selectivity coefficient
depending on the nature of the sensing components in the
membrane. In the case of an ion exchanger, the ISE potential is
controlled by an ion-exchange mechanism at the water-membrane
interface. In this case, the selectivity coefficient is directly related
to the ion-exchange constant of the primary ion and the interfering
ion, KIJ, and it is a function of ion lipophilicity. In the case of a
neutral carrier, the presence of a neutral ionophore with a
relatively large complex formation constant to a specific ion
along with lipophilic ionic sites alters the electrode selectivity
pattern from the normal lipophilicity dependence. The selectiv-
ity coefficient depends on the stability constants of the primary
ion, ꢀIL , and the interfering ion, ꢀJL , with the ionophore, in
corresponding electrode showed an inverse relation-
pot
+,K+
ship between log K
and temperature. Additionally,
Na
sodium-selective electrodes using monensin, mon-
ensin methyl ester (MME), and monensin decyl ester
(MDE) were evaluated as models of electrodes based
on neutral carriers. The electrode based on monensin
showed an enhancement in the selectivity for sodium
over potassium by half an order of magnitude as the
temperature increased from 20 to 50 °C, while the
electrodes based on MME or MDE showed a very small
change in selectivity. This can be explained in light of
changes in the formation constants between the ions
and ionophores with temperature. The theory of the
effect of temperature on selectivity for both ion ex-
changers and neutral carrier ionophores is also
discussed.
nI
nJ
addition to the ion-exchange constant, KIJ. Because the selectiv-
ity coefficient is related to these thermodynamic parameters,
it is expected that it would demonstrate a dependence on
temperature.
The effect of temperature on the ISE’s characteristics is evident
in the Nernst equation;17 the electrode slope increases theoreti-
cally by ∼2 mV/decade with each 10 °C temperature change.
Nearly all the work in the literature has been directed at studying
the effect of temperature on the electrode slope and determining
Improving the selectivity and detection limit of ion-selective
electrodes (ISEs) has been the focus of recent research in the
field of potentiometric sensors. Several strategies have been
explored to enhance the selectivity of ion-selective electrodes.1-10
These mainly include the design of new ionophores with unique
(8) Sokalski, T.; Ceresa, A.; Zwickl, T.; Pretsch, E. J. Am. Chem. Soc. 1997,
* Corresponding author: (e-mail) bachas@uky.edu; (phone) (859) 257-6350;
(fax) (859) 323-1069.
119, 11347–11350
(9) Lee, K. S.; Shin, J. H.; Han, H. H.; Cha, G. S. Anal. Chem. 1993, 65, 3151–
3155
(10) Gemene, K. L.; Shvarev, A.; Bakker, E. Anal. Chim. Acta 2007, 583, 190–
196
(11) Nicolsky, B. P. Zh. Fiz. Khim. 1937, 10, 495
(12) Sandblom, J.; Eisenman, G.; Walker, J. L. J. Phys. Chem. 1967, 71, 3871–
3878
.
† University of Kentucky.
.
‡ Current address: Medicon Hellas, Gerakas 15344, Greece.
§ Universitat Auto`noma de Barcelona.
.
(1) Bader, I. H. A.; Diaz, M.; Hawthorne, M. F.; Bachas, L. G. Anal. Chem.
.
1999, 71, 1371–1377
(2) Johnson, R. D.; Pinchart, A; Bader, I. H. A.; Gingras, M.; Bachas, L. G.
Electroanalysis 2002, 14, 1419–1425
(3) Hassan, S. S. M.; Abou Ghalia, M. H.; Amer, A. E.; Mohamed, A. H. K.
Talanta 2003, 60, 81–91
(4) Gyurcsanyi, R. E.; Lindner, E. Anal. Chem. 2002, 74, 4060–4068
(5) Ammann, D.; Pretsch, E.; Simon, W.; Lindner, E.; Bezegh, A.; Pungor, E.
Anal. Chim. Acta 1985, 171, 119–29
(6) Rosatzin, T.; Bakker, E.; Suzuki, K.; Simon, W. Anal. Chim. Acta 1993,
280, 197–208
.
.
.
(13) Morf, W. E. The Principles of Ion-Selective Electrodes and Membrane
Transport; Elsevier: New York, 1981.
(14) Bakker, E.; Meruva, R. K.; Pretsch, E.; Meyerhoff, E. Anal. Chem. 1994,
.
.
66, 3021–3030
(15) Yoshida, Y.; Matsui, M.; Maeda, K.; Kihara, S. Anal. Chim. Acta 1998,
374, 269–281
(16) Bakker, E.; Pretsch, E.; Bu¨hlmann, P. Anal. Chem. 2000, 72, 1127–1133
.
.
.
.
.
(17) Lindner, E.; Toth, K.; Pungor, E. Dynamic Characteristics of Ion-Selective
(7) Johnson, R. D.; Bachas, L. G. Anal. Bioanal. Chem. 2003, 376, 328–341
.
Electrodes; CRC Press: Boca Raton, FL, 1988.
3622 Analytical Chemistry, Vol. 82, No. 9, May 1, 2010
10.1021/ac902867d  2010 American Chemical Society
Published on Web 04/05/2010

the usable temperature range of the electrode.18-20 Recently, a
new approach to modulate the response of the ISEs has been
introduced.21 In this method, an implanted integrated circuit was
used to heat locally the polymeric membrane through a high
temperature pulse. This temperature pulsing enhances the elec-
trode slope and improves the detection limits.
the Nicolsky-Eisenman equation are not meaningful for interfer-
ing ions that do not demonstrate a Nernstian response.15 Likewise,
when two ions of different charge numbers contribute to the
electrode potential, the magnitude of the selectivity coefficient
differs significantly from that of ions of the same charges. The
widespread use of ISEs, especially in clinical and environmental
Temperature-dependent selectivity has been demonstrated
previously with ion-exchange resins, but has yet to be explored
in the field of ion-selective electrodes. For example, Valiente and
co-workers found a strong dependence of the selectivity of
iminodiacetic acid (IDA) on temperature;22 they used a selective
thermostriping technique to separate copper from zinc at 60 °C.
Muhammed and co-workers23 found a strong dependence of the
loading capacity of chelating resins based on iminodiacetic acid,
picolinic acid, and thiourea on temperature. They used that
property to separate selectively Zn2+ and Cu2+. In another study,
the binding efficiencies of lead and cadmium to Amberlite IR-
120 were shown to have a strong dependence on the temper-
ature of the medium;24 this dependence helped in the removal
of these toxic metals from natural water. These temperature-
dependent selectivity results inspired us to study the effect of
temperature on the selectivity of ISEs based on both charged
and neutral carrier ionophores.
In this study, we investigated the effect of temperature on the
selectivity pattern of two types of ISEs. Electrodes based on
potassium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (KTFPB)
were studied as a model of an ion exchanger based electrode.
Further, monensin ionophores were used to prepare neutral
carrier based electrodes. The theory of the effect of temperature
on the selectivity of both types of electrodes is also discussed.
This study is different from prior work that evaluated the effect
of temperature on the slope of an ISE’s calibration plot18 or
determined the usable temperature range of the electrode.19,20
analysis, requires both high selectivity and accurate knowledge
pot
of K values. The effect of temperature on the selectivity
IJ
coefficient depends on the response of the ISE, and it is
discussed below for the cases of cation-selective electrodes that
incorporate ion exchangers or neutral ionophores.
Ion Exchanger. In the case of membranes that contain an
ion exchanger, the response of the electrode is mainly based on
the relative extraction of the ions into the membrane phase, and
the selectivity coefficient is deduced from the phase boundary
potential:
aI(aq)
aI(org)
RT
zF
RT
zF
EPB
)
ln kI +
ln
(2)
where kI is the so-called single ion distribution coefficient and
is a function of the relative free energies of solvation in both
the sample and membrane phases; kI depends on the lipophi-
licity of the ion. The parameters aI(aq) and aI(org) are the activities
of the free primary ion in the aqueous sample and membrane
phase, respectively.
µ°
( )
- µ°
I org
RT
(
)
I
aq
ln kI )
(3)
where µI(aq)
°
and µ°I(org) are the chemical potentials at the standard
state of the primary ion in the aqueous sample and membrane
phase, respectively. According to eq 9 in ref 14, the ion-
exchange equilibrium constant is given by
THEORY
The selectivity of ion-selective electrodes has been described
by the empirical Nicolsky-Eisenman equation:
kJz /z
aI
I
J
[JzJ]
aJ
pot
IJ
K
) KIJ
)
)
(4)
(
[IzI])(
)
kI
RT
zF
I
J
E ) EI° +
ln(a +
KpotaJz /z
)
(1)
∑
I
IJ
where KIJ is the ion-exchange equilibrium constant. For ions
of the same charge number, if kI and kJ demonstrate different
dependences on temperature, the temperature dependence of
KIJ can be deduced from the van’t Hoff equation
This equation received criticism in recent interpretations of the
true thermodynamic meaning of the selectivity coefficient.16,25-27
It was shown that the selectivity coefficients calculated by using
(18) Shoukry, A. F.; Shuaib, N. M.; Ibrahim, Y. A.; Malhas, R. N. Electroanalysis
d ln KIJ
dT
∆H°
ex
RT2
2005, 17, 713–718
(19) Aghaie, H.; Giahi, M.; Monajjemi, M.; Arvand, M.; Nafissi, G. H.; Aghaie,
M. Sens. Actuators, B 2005, 107, 756–761
(20) Teresa, M.; Vasconcelos, S. D.; Machado, A. S. C. Analyst 1990, 115, 195–
200
.
)
(5)
.
to be as follows:
.
(21) Chumbimuni-Torres, K. Y.; Thammakhet, C.; Galik, M.; Calvo-Marzal, P.;
Wu, J.; Bakker, E.; Flechsig, G.-U.; Wang, J. Anal. Chem. 2009, 81, 10290–
pot
10294
(22) Muraviev, D.; Gonzalo, A.; Valiente, M. Anal. Chem. 1995, 67, 3028–3035
(23) Zagorodni, A. A.; Muiraviev, D. N.; Muhammed, M. Sep. Sci. Technol. 1997,
32, 413–429
(24) Cerjan-Stefanovic, S.; Blanusa, M.; Kastelan-Macan, M. Fresenius’ Z. Anal.
Chem. 1984, 319, 304–305
(25) Bakker, E. Anal. Chim. Acta 1997, 350, 329–340
(26) Meier, P. C.; Morf, M.; Laubli, M.; Simon, W. Anal. Chim. Acta 1984,
156, 1–8
(27) Bakker, E.; Bu¨hlmann, P.; Pretsch, E. Chem. Rev. 1997, 97, 3083–3132
.
K
K
∆H°
R
IJ,T2
ex
1
T2
1
T1
.
ln
) -
-
(6)
pot
IJ,T1
[
]
.
pot
In this case, the plot of ln K vs 1/T should be linear with a
.
IJ
.
slope equal to -(∆H° /R).
ex
Neutral Carrier Ionophore. For cation-selective membrane
electrodes containing neutral carrier ionophores and an optimum
.
.
Analytical Chemistry, Vol. 82, No. 9, May 1, 2010 3623

molar ratio of ionic sites (based on the cation charge number),14,28
the selectivity coefficient depends on the complex formation
constants of each of the primary, ꢀIL , and interfering, ꢀJL , ions
of the temperature in the studied temperature range, eq 13 can
be rewritten by substituting the right-hand side of van’t Hoff’s
equation:
nI
nJ
with the carrier ionophore in the membrane as well as on the
-16
ion exchange constant, KIJ:14
pot
K
K
∆H
∆H
°
°
IJ,T2
ex
1
T1
1
T2
JL
1
T1
1
T2
ln
)
-
+
-
-
pot
IJ,T1
[
]
[
]
R
R
zI
[ILn
]
I
zI
I
IzI + nIL h ILn
ꢀIL
)
)
(7)
(8)
∆H
°
JL
1
T1
1
T2
[IzI][L]nI
n
I
-
(14)
(15)
[
]
R
zJ
J
This can be rearranged to
[JLn
]
zJ
JzJ + nJL h JLn
ꢀJL
[JzJ][L]nJ
J
n
J
pot
K
K
∆H° + ∆H° - ∆H°
IJ,T2
ex
JL
IL
1
T1
1
T2
By considering electroneutrality and mass balance, the theoretical
selectivity coefficient can be calculated from the following equation
as described by Bakker et al.:14,27
ln
)
-
pot
IJ,T1
[
]
R
In practice, this equation implies that, by selecting an appropriate
temperature for the experiment, one may obtain improved
zI/zJ
zI/zJ
J
RT-
zJ[LT - nJ(R-T/zI)]n
ꢀJL
n
J
selectivity for the primary ion. If ∆H°IL ≈ ∆H° , their effects will
pot
IJ
JL
K
) KIJ
zI[LT - nI(R-T/zI)]n
RT-
I
ꢀIL
[
]
cancel each other out, and the change in the selectivity
coefficient with temperature will depend only on the ion-
exchange reaction. On the other hand, when ∆H° and ∆H°
n
I
(9)
IL
JL
are very different, as a result of the selective complexation of
the primary ion, they would affect the selectivity coefficient
with a change in temperature.
If the standard enthalpies of the ion-exchange, complexation
of the interfering ion, and complexation of the primary ion
reactions are dependent on the temperature in the studied
temperature range, the effect of temperature on the selectivity
will be very complex and needs to involve the standard heat
capacity, C°P, of these reactions:
where LT and R-T are the total membrane concentrations of
neutral carrier ionophore and anionic sites, respectively. For
constant concentrations of carrier, LT, and anionic, R-T, sites
and in the case of ions having the same charge (i.e., zI/zJ ) 1),
this equation simplifies to
ꢀ
JLnJ
pot
IJ
K
) KIJ
(10)
ꢀ
ILnI
This equation can be written in the logarithmic form as
T2
∆H° ) ∆H° +
∆C°dT
(16)
∫
T2
T1
P
T1
pot
ln K ) ln KIJ + ln ꢀJL - ln ꢀIL
(11)
IJ
This could lead to a more complex expression of the effect of
n
n
I
J
temperature on the selectivity coefficient, which would result in
pot
By taking the derivative of eq 11 with respect to temperature
a nonlinear plot of ln K vs 1/T; this was not observed to be
IJ
the case in the present study.
pot
IJ,T
d ln ꢀJL
dT
d ln ꢀIL
dT
d ln K
dT
d ln KIJ
dT
n
n
I
J
EXPERIMENTAL SECTION
)
+
-
(12)
Reagents. 2-Nitrophenyl octyl ether (NPOE), dibutyl phthalate
(DBP), poly(vinyl chloride) (PVC), and bis(2-ethylhexyl) sebacate
(DOS) along with calcium chloride were purchased from Fluka
(Ronkonkoma, NY). Sodium chloride and potassium chloride were
obtained from EMD Chemicals (Gibbstown, NJ). Monensin
sodium salt, monensin decyl ester (MDE), and KTFPB along with
the plasticizer dibutyl sebacate (DBS) and tris(hydroxymethy-
l)aminomethane (Tris) were purchased from Sigma (St. Louis,
MO). Lithium chloride was purchased from J. T. Baker (Phillips-
burg, NJ). Monensin methyl ester (MME) was purchased from
Calbiochem (San Diego, CA). Monensin (Mon-H) was obtained
from its commercial sodium salt by dissolving 100 mg in 10 mL
of 0.1 M HCl, followed by extraction with two 5 mL volumes of
chloroform. The chloroform was evaporated overnight in a vacuum
oven at 40 °C, leaving a white powder of monensin acid. Deionized
water of 18 MΩ (Milli-Q water purification system; Millipore,
Bedford, MA) was used throughout the experiments.
The values of the selectivity coefficients at different temperatures
can be determined from the integrated form of eq 12:
pot
ꢀ
ꢀ
ꢀ
ꢀ
K
K
K
K
JLnJ,T2
ILnI,T2
IJ,T2
IJ,T2
ln
) ln
+ ln
- ln
(13)
pot
IJ,T1
IJ,T1
JLnJ,T1
ILnI,T1
It is evident from eq 13 that the temperature dependence of the
selectivity coefficient of neutral carrier based electrodes can be
deduced from the temperature dependence of the three thermody-
namic parameters KIJ, ꢀJL , and ꢀIL . By assuming that the standard
nJ
nI
enthalpy of the ion-exchange, complexation of the interfering ion,
and complexation of the primary ion reactions are independent
(28) Amemiya, S.; Buehlmann, P.; Pretsch, E.; Rusterholz, B.; Umezawa, Y. Anal.
Chem. 2000, 72, 1618–1631
.
3624 Analytical Chemistry, Vol. 82, No. 9, May 1, 2010

Apparatus. The potentiometric measurements were per-
formed using an eight-channel SC-2345 interface (SCC-LP01,
National Instruments, Austin, TX). This interface was connected
to a computer, through a PCI-6036E multifunction I/O data
acquisition device (National Instruments). The computer was
running Lab VIEW 7.0 (National Instruments) software with a
custom-designed program for data acquisition and analysis. The
temperature was controlled using a Fisher Isotemp circulator bath
(model 9500). The reference electrodes used were Orion Ag/AgCl
double-junction electrodes. The pH studies and buffer preparations
were performed with the aid of a Fisher Scientific Accumet 915
pH/ion meter, which was equipped with an Orion pH glass
electrode (Beverly, MA).
Membrane and Electrode Preparation. The membranes
were prepared by dissolving 5.0 mg of the specified ionophore
(corresponding to 2.5 wt %), 25 mol % (relative to the ionophore)
KTFPB, 65 wt % plasticizer o-NPOE, and 32 wt % PVC in 1.0 mL
of THF. These compositions were used for all experiments unless
otherwise stated. The solution was vortexed until the contents
were dissolved and then poured into a 22 mm i.d. glass ring
located on a glass slide, and the solvent was allowed to evaporate
at room temperature overnight. The membrane was cut into
smaller disks, which were mounted onto Philips IS-561 (Gla¨sbla-
serei Mo¨ller, Zurich, Switzerland) electrode bodies.
Potentiometric Measurements. Before use, the membrane
electrodes were conditioned overnight in 0.01 M NaCl prepared
in 0.1 M Tris-HCl buffer, pH 7. The change in the electrode
potential was recorded as an average of three identical electrodes
by placing the ISEs in conjunction with the reference electrode
in a stirred Tris-HCl buffer, pH 7, and adding incremental
amounts of successive concentrations of primary or interfering
ion solutions. The selectivity coefficients were measured by the
separate solution method according to ref 16. All data reported
are average responses of three separate electrodes.
Figure 1. pH dependence of the potentiometric response of the
monensin sodium salt based membrane electrode. pH adjusted by
triethylamine and HCl at (×) pH 2, (0) pH 3, (2) pH 4, and (b) pH 7.
prepared with as-received monensin sodium salt showed a sub-
Nernstian slope of 39.60
0.30 mV/decade for sodium in
Tris-HCl buffer, pH 7; however, they showed a good response
of 54.20 0.50 mV/decade in water. This difference was further
explored by measuring the electrode response of different pH
solutions from pH 2 to pH 7. Figure 1 shows that the electrode
response improves at lower pH; a response for sodium of 58.30
0.60 mV/decade was found at pH 3. This is because the type of
complex formed between monensin and sodium depends on the
degree of protonation of the monensin carboxylic group. Such
dependence on pH was also observed by Suzuki et al.29,30
Monensin forms two sodium complexes of different configurations
and stabilities.31 In neutral and acidic solutions, the sodium
complex is formed with the neutral monensin molecule rather
than the deprotonated monensin.32 This electrode showed an
unusual response with a super-Nernstian slope toward Ba2+ at
pH 7; this interference diminishes at pH 3, which agrees with
the results of Bu¨hlmann et al.33 However, the interference from
lithium and potassium is still present at all pH values.
Effect of Temperature. For studying the effect of temperature
on the electrode potential, the measuring beaker was immersed
in the bath of an isotemperature circulator. Two different types
of experiments were performed to evaluate the effect of temper-
ature on the electrode potential. First, calibration plots for Na+
,
Li+, K+, Ca2+, and Ba2+ were determined at 10, 20, 30, 40, 50,
60, and 70 °C. The electrodes were left to equilibrate for 10
min at each temperature before the potentials were recorded.
Second, the temperature stabilities of the potential and hys-
teresis curves were determined by measuring the potential
changes of a solution containing 1 × 10-3 M Na+ while scanning
the temperature range (10-70 °C).
Monensin was prepared from the commercially available
sodium salt as described in the Experimental Section. As shown
in Figure 2, the sodium response of the corresponding electrode
had a slope of 58.78
0.54 mV/decade. The Li+ and K+
interference still persisted, while the Ba2+ interference was
diminished. The electrodes based on monensin methyl ester
and monensin decyl ester showed a Nernstian sodium response
of 57.60 0.36 and 59.77 0.42 mV/decade for MME and
MDE, respectively, but they still had interference from potas-
sium and lithium ions.
RESULTS AND DISCUSSION
The goal of this study was to evaluate the effect of temperature
on the selectivity of ion-selective electrodes. We used two different
membrane electrode systems. KTFPB was used as the model for
an ion exchanger based electrode. Monensin ionophores were
used for preparing neutral carrier based electrodes.
The electrode based on the ion exchanger, KTFPB, showed a
normal lipophilicity-dependent response (i.e., the electrode was
more selective to larger size cations due to the larger hydration
energies of the smaller ions).
Effect of Temperature on the Response. It is well-known
that the electrode slope changes as a function of temperature. All
electrodes, except the one based on monensin sodium salt,
(29) Tohda, K.; Suzuki, K.; Kosuge, N.; Watanabe, K.; Nagashima, H.; Inoue,
H.; Shirai, T. Anal. Chem. 1990, 62, 936–942
(30) Suzuki, K.; Tohda, K.; Aruga, H.; Matsuzoe, M; Inoue, H.; Shirai, T. Anal.
Chem. 1988, 60, 1714–1721
(31) Gertenbach, P. G.; Popov, A. I. J. Am. Chem. Soc. 1975, 95, 4738–4744
(32) Hoogerheide, J. G.; Popov, A. I. J. Solution Chem. 1979, 8, 83–95
(33) Bu¨hlmann, P.; Amemiya, S.; Umezawa, Y. Anal. Chem. 1998, 70, 445–
454
.
.
.
.
Four different monensin derivatives were used for preparing
the neutral carrier based sodium-selective electrode. Electrodes
.
Analytical Chemistry, Vol. 82, No. 9, May 1, 2010 3625

Figure 4. Effect of temperature on the electrode’s standard potential
(calculated from the extrapolation of the sodium calibration plot) of
(∆) Mon-H, ([) MME, (0) MDE, and (×) KTFPB based membrane
electrodes in 0.1 M Tris-HCl buffer, pH 7.
Figure 2. Potentiometric response of the monensin based mem-
brane electrode in 0.1 M Tris-HCl buffer, pH 7, to (]) Na+, (0) K+,
(2) Li+, (O) Ba2+, and (×) Ca2+
.
1 mV over the temperature range from 10 to 60 °C. Thus, the
change in the cell standard potential is determined by the change
in the ISE standard potential.
The standard potentials were plotted versus temperature; the
slope of this plot gives the isothermal temperature coefficient of
the electrode, ∂E°cell/∂T. As shown in Figure 4, the standard
potentials of Mon-H, MDE, MME, and KTFPB based electrodes
vary linearly with temperature with slopes, ∂E°cell/∂T, of 0.54, 0.43,
0.63, and -0.89 mV/°C, respectively. The electrode based on
monensin sodium salt showed a nonlinear temperature depen-
dence (results not shown), which, as explained earlier, may
be due to the formation of two different complexes of monensin
with Na+.29,30
Effect of Temperature on Selectivity. The main purpose of
this work is to investigate the effect of temperature on the
selectivity of ion-selective electrodes. The selectivity pattern of
each electrode was determined at different temperatures using
the separate solution method described in ref 16. As noted earlier,
the response of the electrode is well-behaved in the range of
20-50 °C, and this range was selected to analyze the effect of
temperature on selectivity. The effect of temperature on the
selectivity of the ion exchanger based membrane electrode is
shown in Figure 5. The logarithm of the selectivity coefficient
showed a linear relationship with the inverse of temperature. This
experimental observation agrees with theory (vide supra) and is
consistent with enthalpy being independent of temperature in the
studied temperature range. The ∆Hex of the ion-exchange
Figure 3. Effect of temperature on the slope of (∆) Mon-H, ([) MME,
and (0) MDE based membrane electrodes in 0.1 M Tris-HCl buffer,
pH 7.
exhibited a near-Nernstian response in the 10-50 °C temperature
range. The isothermal temperature coefficient, d(slope)/dT, of
each of the prepared electrodes was determined by plotting the
electrode slope vs temperature. Figure 3 shows a linear relation-
ship for all the electrodes over the range of 10-50 °C. Slight
deviations from linearity were observed when the temperature
was elevated to 60 °C. This could be attributed to changes in the
gelation properties of the PVC membrane.34 This is also consistent
with prior findings.35,36 Before determination of the effect of
temperature on the standard potential of the prepared electrodes,
the effect of temperature on the double-junction reference
electrode was evaluated by measuring the potential between two
identical Orion double-junction reference electrodes in Tris-HCl
buffer, one used as an indicator electrode in the thermostating
bath and the other used as a reference electrode at room
temperature; the circuit was closed by a salt bridge. It was found
that the change in the reference electrode potential was less than
reaction was determined from the slope of the linear plot of
pot
+,K+
log K
versus 1/T to be -2.53 kJ/mol. The small value of
Na
∆Hex agrees with ∆Hex values determined elsewhere using ion-
exchange resins37 and reveals that the van’t Hoff equation can
describe the temperature response of ion exchanger based
membranes (see eq 6).
The results of the effect of temperature on the selectivity of
MME and MDE based electrodes are listed in Table 1. These
electrodes showed a very small change in the selectivity coef-
pot
+,K+
ficient, log K
, in the temperature range of 20-50 °C that
Na
(34) Aoki, Y.; Li, L.; Kakiuchi, M. Macromolecules 1998, 31, 8117–8123
(35) Arvand, M.; Asadollahzadeh, S. A. Talanta 2008, 75, 1046–1054
(36) Khalil, S.; Abd El-Aliem, S. J. Pharm. Biomed. Anal. 2002, 27, 25–29
.
.
(37) Kolpachnikova, M. G.; Penner, N. A.; Nesterenko, P. N. J. Chromatogr.
1998, 826, 15–23
.
.
3626 Analytical Chemistry, Vol. 82, No. 9, May 1, 2010

Figure 5. Effect of temperature on the sodium selectivity of the
KTFPB based membrane electrode (1 wt % KTFPB) in 0.1 M
Tris-HCl buffer, pH 7.
Figure 6. Effect of temperature on the sodium selectivity of the
monensin acid based membrane electrode in 0.1 M Tris-HCl buffer,
pH 7.
Table 1. Logarithmic Selectivity Coefficients, log
pot
+,K+
K
, Determined by the Separate Solution Method
Na
for Various Sodium-Selective Membrane Electrodes at
Different Temperaturesa
temp,
°C
Mon-H log
MME log
MDE log
KTFPB log
pot
pot
pot
pot
+,K+
+,K+
Na
+,K+
Na
+,K+
K
Na
K
K
K
Na
20
30
40
50
-0.72 0.03
-0.98 0.05
-1.05 0.02
-1.12 0.01
-0.91 0.02
-0.87 0.03
-0.87 0.04
-0.94 0.01
-0.95 0.05
-0.99 0.01
-0.96 0.03
-1.02 0.04
2.21 0.06
2.52 0.03
2.69 0.07
2.86 0.04
a Data shown are averages standard deviation, n ) 3.
was less than 0.07. This value is on the same order of
magnitude as the corresponding standard deviation of log
pot
+,K+
K
, which is 0.03. A possible explanation for these small
Na
changes, in light of eq 15, is that for ionophores MME and MDE
the numerator ∆Hex + ∆H° - ∆H° has a small value.
°
JL
IL
Therefore, the change in temperature results in a very small
change in the selectivity pattern in the studied temperature
range. On the contrary, as is evident from Figure 6, the
selectivity of sodium over potassium of the monensin based
electrode improved by about half an order of magnitude on
elevation of the temperature from 20 to 50 °C.
Why is there a temperature dependence of selectivity for the
neutral carrier based electrodes? This temperature-dependent
selectivity can be explained by employing the theoretical model
described above and assuming that the complexation reaction of
monensin with the alkali-metal cations has a relatively high
dependence on temperature. The complexation constants of
monensin with alkali-metal cations can be used to support this
hypothesis. Popov and Hoogerheide found that the complexation
reaction of monensin and sodium in methanol is enthalpy-
stabilized and entropy-destabilized and the logarithm of the
complexation constant demonstrated an inverse relationship with
temperature.32,38 Because the complexation of monensin and
alkali-metal cations involves formation of a pseudocyclic structure
stabilized by intramolecular hydrogen bonding between both
oxygens of the carboxyl group and two hydroxyl groups at the
Figure 7. (A) The monensin carboxyl group is involved in head-to-
tail hydrogen bonding with the two terminal hydroxyl groups, forming
a pseudocyclic structure. These hydrogen bonds remain upon
complexation of sodium ion. (B) Weak bifurcated hydrogen bonding
of the monensin ester carbonyl group with the two terminal hydroxyl
groups. This hydrogen bond is broken upon complexation to sodium
ion.
other end of the molecule (Figure 7A),39-41 the difference in the
size of the cations enhances the monensin selectivity for sodium
over potassium. Furthermore, Walba and Hermsmeier found that
the conformational energy of monensin is raised in the K+
complex relative to the Na+ complex.42 This increase in
conformational energy favors the smaller size cation with
increasing temperature.
(39) Duax, W. L.; Smith, D. G.; Strong, P. D. J. Am. Chem. Soc. 1980, 102,
6725–6729
(40) Ward, D. L.; Wei, K.; Hoogerheide, J. G.; Popov, A. I. Acta Crystallogr. 1978,
B34, 110–115
(41) Przybylski, P.; Huczyn´ski, A.; Brzezinski, B. J. Mol. Struct. 2007, 826, 156–
164
(42) Walba, D.; Hermsmeier, M. Chem. Commun. 1985, 383–384
.
.
.
(38) Hoogerheide, J. G.; Popov, A. I. J. Solution Chem. 1978, 7, 357–372
.
.
Analytical Chemistry, Vol. 82, No. 9, May 1, 2010 3627

Why is there a difference in the behavior of the monensin ester
based electrodes vs those based on monensin itself? This could
be explained by the different structures of the ion-ionophore
complexes in the two types of ionophores. It has been demon-
strated that monensin behaves as a neutral carrier in polymer
membrane based ISEs; the carboxyl group of monensin does not
become deprotonated upon complexation with Na+ in the mem-
CONCLUSION
The effect of temperature on the selectivity of PVC membrane
ion-selective electrodes was evaluated using both theory and
experiment. The ion exchanger, KTFPB, was found to fulfill the
van’t Hoff equation by showing a linear relationship between log
pot
K
and the reciprocal of temperature. On the other hand, the
IJ
neutral carriers (monensin ionophores) showed two different
behaviors. Monensin acid showed an enhancement of the
selectivity with increasing temperature to 50 °C, while mon-
ensin decyl ester and monensin methyl ester showed an
insignificant change in the same temperature range. Both
effects were explained by theory in light of different contribu-
tions of the ion-ionophore stability constant to the effect.
Performing measurements at different temperatures as outlined
in this paper could be applied to other ionophores with known
high enthalpic complexation dependence to tune the selectivity
of the corresponding ion-selective electrodes.
,30
brane environment.29 As noted above, the carboxyl group is
involved in intramolecular hydrogen bonding with two terminal
hydroxyl groups, resulting in the formation of a pseudomacrocycle
(Figure 7A), which causes preorganization of the structure and
stabilizes the encapsulated cation. On the other hand, in the case
of a monensin ester derivative, Huczynski et al. have shown that
the oxygen atom of the CdO group is engaged in very weak
bifurcated intramolecular hydrogen bonds with the two OH groups
in the free ionophore (Figure 7B). These hydrogen bonds are
broken upon formation of a metal complex.43,44 This intrinsic
difference in the structures of the ion-ionophore complexes
between monensin and monensin ester could account for the
observed difference in the temperature dependence of selectivity.
It should be noted that the strength of hydrogen bonds is
exquisitely sensitive to temperature.45
ACKNOWLEDGMENT
We acknowledge support from the National Aeronautics and
Space Administration (L.G.B.) and the Ministry of Higher Educa-
tion, Egypt (E.M.Z.). The Spanish Ministerio de Ciencia e
Innovacio´n is acknowledged for partially supporting the present
study (Project CTM2006-13091-C02-02/TECNO) (M.V.).
(43) Huczyn´ski, A.; Przybylski, P.; Brzezinski, B. Tetrahedron 2007, 63, 8831–
8839
(44) Huczyn´ski, A.; Przybylski, P.; Brzezinski, B.; Bartl, F. Biopolymers 2006,
81, 282–294
(45) Dougherty, R. C. J. Chem. Phys. 1998, 109, 7372–7378
.
Received for review December 16, 2009. Accepted March
11, 2010.
.
.
AC902867D
3628 Analytical Chemistry, Vol. 82, No. 9, May 1, 2010